Abstract
Texture modeling and analysis are challenging issues of image processing. In many cases, the model has to incorporate some important characteristics of the data as roughness or anisotropy properties, that can be handled using a stochastic approach, involving fractional anisotropic random fields. We give a sense to the notion of orientation for self-similar Gaussian fields with stationary increments, based on a Riesz analysis of these fields, with isotropic zero-mean analysis functions. We propose a structure tensor formulation and provide an intrinsic definition of the orientation vector as eigenvector of this tensor. That is, we show that the orientation vector does not depend on the analysis function, but only on the anisotropy encoded in the spectral density of the field. Then, we generalize this definition to a larger class of random fields called localizable Gaussian fields, whose orientation is derived from the orientation of their tangent fields. Finally two classes of Gaussian models with prescribed orientation are studied in the light of these new analysis tools.
Link to the corresponding article.